Innovative approach for engineering NLG systems: the content determination case study

  • Authors:
  • Marco Fonseca;Leonardo Junior;Alexandre Melo;Hendrik Macedo

  • Affiliations:
  • Departamento de Computação, Universidade Federal de Sergipe, São Cristóvão, SE, Brazil;Departamento de Computação, Universidade Federal de Sergipe, São Cristóvão, SE, Brazil;Departamento de Computação, Universidade Federal de Sergipe, São Cristóvão, SE, Brazil;Departamento de Computação, Universidade Federal de Sergipe, São Cristóvão, SE, Brazil

  • Venue:
  • CICLing'08 Proceedings of the 9th international conference on Computational linguistics and intelligent text processing
  • Year:
  • 2008

Quantified Score

Hi-index 0.00

Visualization

Abstract

The purpose of Natural Language Generation (NLG) systems is that of automating the production of linguistically correct texts from a data source. Generators are usually built using ad-hoc software engineering practices, lacking a well-defined development process, standard software architecture, and the use of worldwide programming languages. This paper describes a new development approach that leverages the most recent programming languages and standards of modern software engineering to enhance the practical use of NLG applications. To show the practicability of the proposal, a content determination system, which accepts as input wrapped Web data regarding soccer championship results, was developed.